An Expedient Mind: PREFACE

AN EXPEDIENT MIND

Tartan Hen Publications : Home | more books | Contact : feedback@tartanhen.co.uk

ZOMBIE-ISM

Every now and then, you come across a killer argument - that is, an argument which brooks no rational denial. There are also killer observations. It is often pointed out that in science there is never 100 percent confidence that any particular explanation or theory is true. Always there is a possibility, however, remote, that some alternative theory might come forward to usurp the position of the current version of received wisdom. And that space for doubt, no matter how small the space may be, is often used my the mischievously minded faithful to cast doubt on whatever scientific theory may contradict their own dearly held beliefs. Not only do they seem unable to recognise the variable size of that doubtful gap, and consistently avert their gaze from the infinitely wide gap in their own credibility, but they always seem to overlook this fact - It may be that we can never be certain that a particular theory is correct, but we can on accasions, be absolutely, 100 percent certain, that a particular theory is wrong.
And that is where the killer arguments and killer observations come in. They finger, with unerring precision, the theories which are wrong - totally and undeniably wrong.

Some famous killer observations -

(1) Galileo pointed a telescope at Jupiter and observed that it has moons which revolve around Jupiter, not around the Earth as was claimed by the Geo-centric Earth theory.

(2) Michaelson and Morley measured the speed of light over a measured distance while the Earth sped round the Sun and turned on its axis. They observed that the speed of light did not shift during that time by so much as a smidgeon, despite the fact that at one moment it was travelling in the same direction as the Earth (round the Sun) and at other times in the opposite direction, or at right angles. And that demolished the idea that light travels through some kind of fixed medium (called the ether) as sound-waves travel through the air.

And some killer arguments -

(3) Back to Galileo and this time to his demonstration that (ignoring air resistance) solid objects fall with the same speed and accelleration. Previously it had been assumed (because it seemed obvious) that large heavy objects fall faster than small ones - just because they are larger. Not only did Galileo do the requisite experiment, but there is also a killer argument to support his findings, and which may well have prompted his experimentation. Consider two solid objects of exactly the same size and shape falling together. Imagine a tiny demon living in the space between them, who, halfway down, decides to tie the two objects together with a piece of thread. Instead of two objects, we now have one single but larger object. Why would it suddenly start to fall faster?

(4) Richard Dawkins, in his book "The God Delusion" offered an argument which kills the notion of Intelligent Design. If it is the case that everything which is complicated must have been designed by an intelligent designer, it must also be the case that that designer is even more complicated than the thing he (or presumably she) designed. Therefore, according to the same rule - who designed the designer? And who designed him/her? And so on ad infinitum. Note that if anyone tries to counter that by supposing that the designer has magic properties which free him/her from the normal rules of complicated things needing a designer, then they have destroyed the claim that the Intelligent design argument is a scientific theory rather than a religious faith and so it has no legitimate claim on a school science syllabus.

The Zombie Argument

And so I come to my own particular silver bullet. Ironically, it is based on an argument, or thought experiement, which was supposed to operate in the reverse direction. For want of a better term, I'll call them the phenomenalists. These are the people who believe that consciousness is either a phenomenon, which is completely non-physical, or that it is a physical phenomenon of a different kind - something, rather vaguely, to do with the peculiarities of quantum mechanics perhaps. All phenomenalists concede that consciousness is related in some way to the physical mechanisms of the brain, but this linkage is generally described as a "correlation", or an "association". The consequence of this idea, is that that linkage can, in theory, be broken. In that case, if would therefore be possible to have the physical mechanism without the attendant phenomenon of consciousness.

ZOMBIES

This state of affairs leads the phenomenalists to suggest the possible existence of zombies. A zombie is a thing, which has all the behavioural characteristics of a human being, and, so far as external observers are concerned, is completely indestinguishable from a human being. The difference is apparent only to the zombie itself, because it is internally, quite unconscious of its own behaviour.
If the phenomenalists could prove that such a thing as a zombie can exist, then they will have proved their case. They would have a killer argument which destroys the physicalist position that consciousness is not simply correlated to the brain mechanism - it IS a brain mechanism. Note, however, that the physicalist case does not claim that being conscious is a property of the whole brain mechanism. It claims only that consciousness is a component of the brain mechanism. So it does not help the phenomenalist claim to point out that we humans are perfectly capable of carrying out certain mental operations without being conscious of what we are doing. It is only when we are doing certain parts of the brain process, that we are conscious. I can put that another way - it is only when we are conscious that we are conscious of what we are doing. That may seem a curiously self-defining, or tautalogical way of expressing it, but it captures the point I am trying to make. Being conscious, in the physical account, means doing a particular kind of procedural operation, in which the brain models and stores for future reference, a model of what it is, itself, doing at that moment. To be conscious, therefore, implies that one is processing information about oneself.
To counter that idea, phenomenalists point out that it is perfectly possible to "process information" about what is going on in the brain without being conscious of being the person whose brain is being discussed. That, I suggest, is a disingenuous argument.
An analogy migght help elucidate. If I walk into a shop and use my credit card to make a purchase, the credit card system invokes a procedure which can be called the "transaction procedure". I can also, at another time, phone up my credit card company and discuss that procedure. But that discussion does not correspond to a "transaction." I can do a transaction only by offering my card in a shop or giving its number to an online website. I actually have to conduct a transaction in order to invoke that procedure.
Likewise, I can discuss brain procedures as much as I want but I cannot actually invoke the consciousness procedure and therefore be conscious in the conventional sense, if I am not actually doing that particular consciousness procedure. As I have tried to show in this book, there are many different ways in which a brain mechanism can process a particular set of information, but only certain of these ways can be called the "consciousness procedure".
David Chalmers is a prominent supporter of the phenomenalist case, and a vociferous proponent of the zombie argument. He bases his version of the argument on what he claims is the "conceivability" of the zombie system. To quote his words verbatim -

(1) It is conceivable that there be zombies.

(2) If it is conceivable that there be zombies, it is metaphysically possible that there be zombies.

(3) If it is metaphysically possible that there be zombies, then consciousness is nonphysical.

(4) Consciousness is nonphysical.

[Chamlmers 2002]

I have to confess to being totally flummoxed by this argument. I do not understand the difference between something being "possible" and its being "metaphysically possible" and I certainly reject the idea that because someone can conceive of something, that that something must be possible. I can, for example, conceive of circular triangles. I can conceive of them, but I cannot imagine how they could be constructed. Without some evidence to the contrary, it could easily be the case that a person who considers some concept to be possible, has simply overlooked some logical contradiction which would render the idea impossible. And that, as I propose to show here, is exactly what has happened here. The philosopher David Lewis, has suggested that we should take the notion of "possible universes" seriously, but I am not persuaded that it has any validity. I regard possible circumstances as just another mental structure, like the structure I call expedient {REALITY} - one which does not appear to be logical contradictory of any circumstances which are part of {REALITY}. The emphasis being on the word "appear".

THE ZOMBIE KILLER ARGUMENT

My argument takes a classic form. It assumes that the zombie argument is sound, and then shows that that leads to a logical contradiction.
So we start by assuming that we have created a zombie. We have constructed a robotic device which is outwardly indesinguishable from a human being in every way. But because it is a man-made mechanical device, it cannot be in possession of that ellusive and mysterious nonphysical consciousness phenomenon.
Being indestinguishable from a human being, however, means that it does not only pass the simple Turin tests, which are (typicall) envisaged by the zombie proponents. It has to pass a much more vigorous test which we will devise for the purposes of this argument. Note that, according to the physicalist account, consciousness is involved only when the mechanism has activated those special parts of its mechanism which process information about its own brain procedures. So our special Turing Test will concentrate on those particular aspects of its thinking. We will ask it questions about observations it is making right now, and has made in the past. "When you were in that room next door, did you see a book lying on the table?" Obviously it can produce a set pattern of responses, which might be adequate for a superficial test, but not on this occasion. We are going to be hard to convince. It cannot say, for example, "Sorry I can't remember" because we will then say, "Go back into the room now, then come out again and tell me if you saw a book on the table." In these circumstances, if it is going to convince us that it is actually conscious, it will need to be able to store a memory trace of what it saw in the room next door, so that it can refer to that before it replies.
And that is a general principle which I want to establish. This zombie has to have internal stores of information to which it can refer and which is constantly, and automatically updated by information coming from its perceptual apparatus.
Next we ask it questions about what it has already told us. So it has to remember its own words, and how it answered previous questions. And why it answered that way.
That's important. We ask it why it gave us a particular answer on a previous occasion. To pass the test it needs to give us an answer which is not only plausible, but compatible with its other pronouncements. In retrospect, it has to understand its own motivation for giving these responses. This is a hard, searching test.
And finally we ask it about its own feeling of being conscious. To convince us that it is speaking the truth - and therefore passing muster as a system indestinguishable from a human being, it has to tell us that it does think it is conscious. Not only that, it has to have internal stores of information to which it can refer, and which tells it that it is indeed conscious. For how else can it define the meaning of the word "conscious" except in terms of its own procedures for accessing the mental store of information it has as a model of its own brain?
It follows, that in order to be indistinguishable from a human being, the zombie has to have a brain mechanism which convinces itself that it is conscious. So what we then have is an unconscious zombie which honestly believes itself to be conscious. Since the accepted criterion of consciousness is the internal experience of being conscious, that is a logic contradiction.

Conclusion

Zombie-ism is impossible, in this, or any conceivable universe. The Turing Test, provided it is rigorously applied, is an adequate test of consciousness.

Postscript
The nature of that zombie-killer argument reminds me of an observation made by Professor R.V.Jones.
During the second world war, Jones played a key role in what became known as "The Battle of the Beams". In his book, "Most Secret War" he tells how the Nazis constructed two powerful high frequency transmitters, one in occupied France and the other in occupied Norway. These transmitters broadcast narrow radio-beams to intersect over some selected British city. The Nazi bombers then flew at night, along one of the beams, until they encountered the other. At that point of intersection, they dropped their bomb load. It was the task of Jones and his colleagues, to defeat these efforts, and in this task, they were largely successful. They jammed the German transmissions and they transmitted false beams designed to encourage the dropping of bombs in empty countryside. These counter-measures, and the German counter-counter-measures, became increasingly sophisticated.
Later, in an essay, written when he was a Professor of Physics at Aberdeen University, Jones reflected on the nature of deception in general. He noted that their efforts to deceive the German bomber crews were bedevilled by the fact that the fake beam transmitters were located in Britain and not in Nazi-occupied Europe. That difference, was always, in principle, detectable. He concluded that in order to produce false beams which would appear to be transmitted from Nazi-held territory (without there being any possibility of the deception being detected), the only place those fake beam transmitters could be constructed, would be actually on Nazi-held territory. This, he reflected, was an unavoidable feature of every deceptive model - and an inherent irony. An undetectable deception, had to be real.
It seems to me that that principle is relevant to my zombie argument. A zombie, to meet its design specification, must put on a fake performance as a human being which is deeply convincing. It must therefore put on a fake performance of being conscious. And to do that, in a way that is undetectable, it has to be really conscious.

Tartan Hen Publications : Home | more books | Contact : feedback@tartanhen.co.uk